Conditional migration and data filters

A filter copies only the rows of a table that match a condition, not the whole table. Use it for large tables, partial exports, or loads bounded by a date.

Build a filter

1. On Tables & mapping, click the table in the list.
2. Click Filter rows… at the top right of the table.
3. Click + Add condition. Pick the column and the operator, and type the value.
4. Add more conditions. Rows must match all joins them with AND; any joins them with OR.
5. Click Refresh to see the rows that match, then OK.

The line under the conditions shows the query the filter makes. Operators: =, <>, >, <, >=, <=, LIKE, NOT LIKE, IS NULL, IS NOT NULL. For LIKE, % matches any text, for example %smith%.

Type values as they are, without quotes: 5, 2005-08-01, Germany. The filter quotes text and dates itself.

Write the condition in SQL

Click Edit as SQL to type the WHERE condition yourself. Here values are written as the source database expects them:

• MySQL / MariaDB, SQL Server, PostgreSQL, Oracle, SQLite: dates and text in single quotes, payment_date >= '2005-08-01'.
• MS Access: dates between # signs, #2005-08-01#.

Back to the condition list returns to the conditions.

Limit rows, or filter every table

• Copy at most copies only the first rows the source returns, in no set order. Leave it empty to copy all that match.
• Apply this filter to all tables sets the same condition on every selected table. It is checked against this table only: a table without these columns fails at the run.

A filter also cuts transfer time on large tables. See Speeding up large transfers.